U.S. patent number 6,945,626 [Application Number 10/656,132] was granted by the patent office on 2005-09-20 for correction table generation method and method of controlling correction table generation apparatus.
This patent grant is currently assigned to Canon Kabushiki Kaisha. Invention is credited to Shigeyasu Nagoshi, Akihiko Nakatani, Okinori Tsuchiya.
United States Patent |
6,945,626 |
Tsuchiya , et al. |
September 20, 2005 |
**Please see images for:
( Certificate of Correction ) ** |
Correction table generation method and method of controlling
correction table generation apparatus
Abstract
In a multi-valued printer that uses discontinuous index
patterns, if a tone correction table is generated using sampled
density patches, a table which is different from a table to be
obtained and has no inflection point is obtained, and the print
density characteristics after tone correction suffer discontinuity.
To solve this problem, an output gamma table used to output
measurement patches is set to linearly correct the printer print
characteristics. Patches are output and their densities are
measured. A reverse table of a "signal value--density" table is
generated, and is smoothed using a recursive curve. The smoothed
reverse table is finely adjusted to generate an intermediate output
gamma table. The generated table undergoes index component
correction, thus generating a tone correction table.
Inventors: |
Tsuchiya; Okinori (Kanagawa,
JP), Nagoshi; Shigeyasu (Kanagawa, JP),
Nakatani; Akihiko (Kanagawa, JP) |
Assignee: |
Canon Kabushiki Kaisha (Tokyo,
JP)
|
Family
ID: |
32396234 |
Appl.
No.: |
10/656,132 |
Filed: |
September 8, 2003 |
Foreign Application Priority Data
|
|
|
|
|
Sep 9, 2002 [JP] |
|
|
2002-263220 |
Sep 9, 2002 [JP] |
|
|
2002-263223 |
|
Current U.S.
Class: |
347/19;
400/74 |
Current CPC
Class: |
H04N
1/40087 (20130101); H04N 1/407 (20130101); H04N
1/6033 (20130101); H04N 1/6097 (20130101) |
Current International
Class: |
B41J
2/205 (20060101); B41J 29/393 (20060101); G06K
15/00 (20060101); B41J 029/393 () |
Field of
Search: |
;347/12,14,19,37,15
;400/74 ;358/406,504 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
2-167755 |
|
Jun 1990 |
|
JP |
|
8-2659 |
|
Jan 1996 |
|
JP |
|
9-46522 |
|
Feb 1997 |
|
JP |
|
Primary Examiner: Stephens; Juanita D.
Attorney, Agent or Firm: Fitzpatrick, Cella, Harper &
Scinto
Claims
What is claimed is:
1. A generation method of generating a correction table used to
correct print characteristics of a printing apparatus which has
nonlinear characteristics between an input signal and a print
signal to be output in accordance with the input signal, said
method comprising step of: generating, on the basis of a first tone
correction table which is generated while compensating for the
nonlinear characteristics, a second tone correction table having a
maximum value different from the first tone correction table.
2. A method according to claim 1, wherein the second tone
correction table is generated on the basis of an input image signal
to the printing apparatus and print density characteristics of the
printing apparatus.
3. A method according to claim 1, wherein the second tone
correction table is generated on the basis of a table, which is
generated from a table corresponding to the nonlinear
characteristics while compensating for the nonlinear
characteristics, to have the first tone correction table as an
input.
4. A method according to claim 1, wherein the second tone
correction table is generated on the basis of a table which is
obtained by variably scaling tone correction values of the first
tone correction table.
5. A method according to claim 1, wherein the second tone
correction table is used to correct a dot quantity of a print
material of a print medium in consideration of an acceptable
quantity of the print material.
6. A control method of generating a correction table used to
correct print characteristics of a printing apparatus which has
nonlinear characteristics between an input signal and a print
signal to be output in accordance with the input signal, said
method comprising steps of: setting a first tone correction table
generated while compensating for the nonlinear characteristics; and
generating a second tone correction table having a maximum value
different from the first tone correction table on the basis of the
first tone correction table.
7. A computer program embodied on a computer-readable storage
medium executing a generation method of generating a correction
table used to correct print characteristics of a printing apparatus
which has nonlinear characteristics between an input signal and a
print signal to be output in accordance with the input signal, said
computer program comprising: code for generating, on the basis of a
first tone correction table which is generated while compensating
for the nonlinear characteristics, a second tone correction table
having a maximum value different from the first tone correction
table.
8. A computer readable medium storing a computer program for
executing a generation method of generating a correction table used
to correct print characteristics of a printing apparatus which has
nonlinear characteristics between an input signal and a print
signal to be output in accordance with the input signal, said
computer program comprising: code for generating, on the basis of a
first tone correction table which is generated while compensating
for the nonlinear characteristics, a second tone correction table
having a maximum value different from the first tone correction
table.
Description
FIELD OF THE INVENTION
The present invention relates to a correction table generation
method, and a method of controlling a correction table generation
apparatus and, more particularly, to generation of a correction
table which is used to correct the print characteristics of a
printing apparatus, which has nonlinear characteristics between an
input signal value and a print signal value output in accordance
with the input signal value.
BACKGROUND OF THE INVENTION
A printing apparatus such as a printer, copying machine, facsimile,
or the like prints an image defined by dot patterns on a print
medium such as a paper sheet, thin plastic plate, or the like on
the basis of image information. Such printing apparatus can be
categorized into an ink-jet system, wire-dot system, thermal
system, laser beam system, and the like in accordance with the
printing systems. Of these systems, the ink-jet system (ink-jet
printing apparatus) prints an image by ejecting ink (print liquid)
droplets from ejection ports of a print head and attaching them on
a print medium.
As an example of a printing apparatus, an ink-jet printing
apparatus (ink-jet printer) with an output resolution of 600 dpi
will be described. This printer prints an image by receiving a
binary (multi-valued) dot image, which is obtained by rasterizing
300-dpi, 8-bit image data and is transferred from a host (PC).
The ink-jet printing apparatus prints an image by combining pixels
(dot ON) formed by print liquid droplets landing on a print medium,
and pixels (dot OFF) without any print liquid droplet.
In recent years, since the ejection ports of a print head can be
arranged at a high density, an image with a relatively high
resolution (e.g., 600 dpi) can be printed. In order to obtain a
high-definition print, when a host as a supply source of image data
connected to a printer processes high-resolution (e.g., 600 dpi)
image data and transfers the processed data to the printer, the
data size becomes four times that required upon transferring
300-dpi image data, and the data processing time and transfer time
increase largely. Hence, a method of handling a binary dot matrix
as a multi-valued super pixel on the printer side (Japanese Patent
Laid-Open No. 9-46522) is used.
[Multi-valued Printer Using Index Pattern]
Japanese Patent Laid-Open No. 9-46522 discloses the following
technique for reducing the loads on the data process and transfer.
With this technique, a binary ink-jet printer with an output
resolution of 600 dpi handles, as one set, 2.times.2 binary pixels
of a resolution of 600 dpi, and a host handles that printer as a
5-valued printer with a resolution of 300 dpi.
FIG. 1 shows a layout method of ON dots when 2.times.2 binary
pixels of a resolution of 600 dpi are used as one set. A set of
2.times.2 pixels will be referred to as a "super pixel", and the
layout of ON dots will be referred to as an "index pattern"
hereinafter.
As shown in FIG. 1, a super pixel has five different index
patterns, which respectively have zero, one, two, three, and four
ON dots. A printer, which holds such index pattern in advance,
prints an image by rasterizing 300-dpi, 5-valued image data input
from the host to 600-dpi, binary image data with reference to the
index patterns.
Pseudo halftoning in the printer using the index patterns will be
explained below.
The host can consider the aforementioned binary printer with the
output resolution of 600 dpi as a 5-valued printer with an input
resolution of 300 dpi, since the printer main body executes a
rasterization process of the index patterns. Hence, the host
rasterizes 300-dpi, 8-bit input image data to a 5-valued dot image
by known pseudo halftoning such as multi-valued error diffusion,
multi-valued dithering, or the like, and supplies the dot image to
the printer.
FIG. 2 shows an average distribution of the number of index
patterns assigned per unit area (ordinate) with respect to the
signal value of an 8-bit input image signal (abscissa). Note that
"average" means that error diffusion and dithering often have
different numbers of locally printed dots at the beginning of
processing of an image to be processed, and such local difference
is not taken into consideration.
As shown in FIG. 2, for example, when the signal value is "0", all
pixels of 300 dpi are occupied by index patterns of index No. 0;
when the signal value is "64", all pixels of 300 dpi are occupied
by index patterns of index No. 1. FIG. 3 summarizes correspondence
between the signal values (central values) and index numbers. Note
that pixels of index Nos. 0 and 1 mix in a region of 0<signal
value<64, and those of index Nos. 1 and 2 mix in a region of
64<signal value<128.
FIG. 4 shows the average number of 600-dpi, binary ON dots per unit
area (ordinate) with respect to the signal value of an 8-bit input
image signal (abscissa).
For example, in a region of 0<signal value<64, the number of
index patterns of index No. 0 decreases, and that of index patterns
of index No. 1 increases. That is, in the region of 0<signal
value<64, the number of ON dots increases with increasing number
of index patterns of index No. 1, as indicated by the solid line in
FIG. 4.
In a region of 64<signal value<128, the number of index
patterns of index No. 1 decreases, and that of index patterns of
index No. 2 increases. Hence, in the region of 64<signal
value<128, since the number of ON dots formed by index patterns
of index No. 1 (600 dpi, one ON dot per four pixels) decreases, and
the number of ON dots formed by index patterns of index No. 2 (600
dpi, two ON dots per four pixels) increases, as indicated by the
broken line in FIG. 4, the total number of ON dots in the region of
64<signal value<128 increases, as indicated by the solid line
in FIG. 4. In this manner, the number of binary ON dots of an
output resolution of 600 dpi, which are formed by the 5-valued
printer with the input resolution of 300 dpi using the index
patterns shown in FIG. 1, increases in proportion to the signal
value of the input image signal.
[Print Characteristics of 5-valued Printer with Input Resolution of
300 dpi]
As described above, the signal value of an image signal input to
the 5-valued printer with the input resolution of 300 dpi is
proportional to the number of ON dots. However, the print density
is normally not proportional to the number of ON dots due to
mechanical and optical dot gains. FIG. 5 shows the print density
(ordinate) upon printing an 8-bit input signal value (abscissa) as
a 600-dpi binary pattern. As shown in FIG. 5, the print density
shows the tendency of leveling off.
[Tone Correction Table Corresponding to Index Pattern]
In order to compensate for the influence of dot gains shown in FIG.
5 and to correct the relationship between the input signal value
and print density to a proportional relationship shown in FIG. 6, a
tone correction table disclosed in Japanese Patent Publication No.
8-2659 (called a "density characteristic correction table" in this
reference) is used.
In the tone correction table, a table shown in FIG. 7, which
defines an inverse function to FIG. 5 used to correct the
relationship between the input signal value and print
characteristics shown in FIG. 5, is set. A 300-dpi, 8-bit image
signal is converted into a tone-corrected image signal by the tone
correction table, and then undergoes pseudo halftoning by, e.g.,
multi-valued error diffusion to be rasterized to a 300-dpi,
5-valued dot image. The rasterized dot image is input to the
printer. The printer rasterizes the 300-dpi, 5-valued dot image to
a 600-dpi, binary dot image with reference to the index patterns.
Then, nozzles of a print head corresponding to ON dots of this
binary dot image are driven to eject print liquid droplets.
FIG. 8 is a flow chart for explaining a process for deriving a tone
correction table for a 5-valued printer with an input resolution of
300 dpi.
The tone correction table is derived by executing a print process
of the printer in correspondence with an input signal value,
measuring the density of a print, and calculating a table which
compensates for the measured density characteristics. Note that
output characteristics of the printer can be examined by measuring
intensity or brightness of the print instead of the density of the
print. In other words, the examination of the output
characteristics of the printer can be performed based on amounts
indicating brightness, such as the intensity, brightness or density
of the print, signal values read from the print by using a scanner,
or the like. In the following explanation, the density is used as
the amounts which quantitatively show the characteristics of the
printer, but the above intensity, brightness or signal values can
be used as the amounts which quantitatively show the
characteristics of the printer, too.
It doesn't matter even if the quantity except that brightness is
shown is substituted for this concentration though concentration is
used as the representative of the quantity which shows brightness
and it explains.
An output gamma used to output measurement patches (a tone
correction table shown in FIG. 9 that allows an input signal value
to pass through it) is set (S102), and measurement patches are set
(S103). The measurement patches are output (S104), and the
densities of the output patches are measured by a densitometer
(S105).
As the measurement patches, those shown in FIG. 10 are used in
consideration of a reduction of a time required from printing to
measurement, and print reproducibility by the ink-jet printer.
Numerals printed above respective patches shown in FIG. 10 indicate
the signal values of the patches, and letters C, M, Y, and K on the
upper end indicate inks used to print patch sequences below these
letters.
A "signal value--density" table is generated based on the patch
density measurement result, and a "signal value--density" table
obtained by normalizing the signal values and density values in the
former table to the range from 0 to 1, as shown in FIG. 11, is
generated. In order to obtain an inverse function of this
normalized table, the normalized table is referred to in the
reverse direction to generate a reverse table of the "signal
value--density" table (S106).
As indicated by .circle-solid. marks in FIG. 11, the measurement
values of actual densities include errors due to reproducibility of
the ink-jet printer upon printing, and measurement errors. Hence,
the reverse table undergoes smoothing based on a recursive curve
using polynomial approximation given by:
to obtain a table that shows a smooth curve, as indicated by the
solid curve in FIG. 11 (S107). Note that a detailed description of
smoothing based on polynomial approximation will be omitted, since
a similar process using a spline curve is disclosed in detail in
Japanese Patent Publication No. 8-2659.
Deviations due to oscillations of the polynomial approximation and
those of an origin (0.0, 0.0) and end point (1.0, 1.0) often occur.
In order to correct those deviations, the reverse table is finely
adjusted (S108). The range of the reverse table, which has been
normalized and finely adjusted in a period from the origin to the
end point, is re-converted to a table format of an 8-bit integer,
thus generating a tone correction table (S109).
By making tone correction using the tone correction table generated
in this way, the tone characteristics can be improved so as to
realize linearity between the input image signal value and output
print density of the printer.
Note that the tone correction table is a linear table for each ink
color, and has a smaller table size than a color conversion table
as a three-dimensional (3D) table. Hence, the following applied
techniques using a plurality of tone correction tables are
available.
Ejection amount correction of print head using tone correction
table
Dot quantity correction using tone correction table
The applied techniques using tone correction tables will be
described in detail below.
[Ejection Amount Correction of Print Head Using Tone Correction
Table]
The output density characteristics of an ink-jet printer depend on
the amount of ink ejected from nozzles of a print head. The
ejection amounts of mass-produced print heads suffers variations of
about .+-.10% with respect to the standard amount due to variations
of the characteristics generated during production.
In general, a color processing unit of a printer is designed to
obtain a desired output result on the basis of the standard
ejection amount of a print head. In other words, if a print head
whose ejection amount deviates from the standard ejection amount is
used, the output result becomes different from the standard output
result expected upon design. That is, the variations of the
ejection amount of the print head cause deterioration of the tone
characteristics such as deterioration of color reproducibility,
unbalance of density inks, and the like.
To solve this problem, Japanese Patent Laid-Open No. 2-167755
discloses the following ejection amount correction method. In this
method, a color processing unit is designed to obtain an optimal
output result when the ejection amount of a print head is a lower
limit value. When a print head which has an ejection amount equal
to or larger than the lower limit value is used, the ejection
amount of the print head used is corrected using a tone correction
table to be equivalent to that of a print head whose ejection
amount is the lower limit value.
FIG. 24 is a block diagram showing the arrangement of a printer
driver 1300 (color processing unit) which runs on a host, and an
ink-jet printer 1400.
R, G, and B image signals (300 dpi, 8 bits per color) input to a
printer driver 1300 are converted into C, M, Y, and K signals (300
dpi, 8 bits per color) by a 3D lookup table (3DLUT) of an RGB/CMYK
converter 1301. Note that a process for a 300-dpi, 8-bit cyan (C)
signal will be explained below, but the same applies to those for
other color component signals.
The 300-dpi, 8-bit (256-valued) cyan (C) signal undergoes tone
correction in a tone correction unit 1302 and is expanded to 12
bits (4081 values). In this case, a tone correction table used by
the tone correction unit 1302 is selected from a tone correction
table database 1305 on the basis of independently input ejection
amount information of the print head. The expanded 300-dpi, 12-bit
C signal undergoes pseudo halftoning by a multi-valued error
diffusion processor 1303 to be converted into a 300-dpi, 3-bit
(5-valued) multi-valued dot image.
The 300-dpi, 3-bit multi-valued dot image is transferred (input)
from the host to a printer 1400. The 300-dpi, 3-bit multi-valued
dot image input to the printer 1400 is rasterized to a 600-dpi,
binary dot image by a dot image rasterize processor 1401, which
refers to index patterns of a super pixel stored in an index
pattern memory 1402.
The 600-dpi, binary dot image is stored in a dot image rasterize
buffer 1403, and is sequentially sent to a print unit 1404 in
correspondence with nozzles arranged in an ink-jet print head at
600-dpi intervals. Print nozzles corresponding to ON dots of the
600-dpi, binary dot image are driven, thus ejecting print liquid
droplets.
A method of deriving the tone correction table used by the tone
correction unit 1302 in ejection amount correction, and the
correction method will be described below. A case will be
exemplified wherein continuous index patterns are used, and print
heads are roughly classified into the following three ranks on the
basis of their ejection amounts. Lower limit value (small ejection
amount): n [ng] Central value (middle ejection amount): 1.05 n [ng]
Upper limit value (large ejection amount): 1.10 n [ng]
When a 12-bit tone correction table is generated using a print head
of the small ejection amount rank, and a maximum value "4080" of
that tone correction table is assigned to a print head of the small
ejection amount rank, the ejection amount is 4080.times.n [ng].
Hence, the maximum values of tone correction tables which can yield
the same ejection amount for the remaining ejection amount ranks
are as follows:
Maximum value of tone correction table of middle ejection
amount=4080.times.n/(1.05.times.n).apprxeq.3886
Maximum value of tone correction table of large ejection
amount=4080.times.n/(1.1.times.n).apprxeq.3709
By controlling to obtain an identical ejection amount after tone
correction using the tone correction table generated for each
ejection amount of a print head, a similar output result can be
obtained independently of the ejection amount ranks of print
heads.
The maximum values of tone correction tables have been explained.
Likewise, when a 12-bit tone correction table is generated using a
print head of the middle ejection amount rank, an identical
ejection amount can be obtained for all input/output image signal
values independently of the ejection amount ranks of print heads by
setting the central values of tone correction tables of the
remaining ejection amount ranks as:
Correction value of large ejection amount rank=correction value of
middle ejection amount rank.times.3709/3886
Correction value of small ejection amount rank=correction value of
middle ejection amount rank.times.4080/3886
A sequence for generating a tone correction table (maximum
value=3886) for a print head of the middle ejection amount rank
using a print head of the middle ejection amount rank will be
described below with reference to the flow chart in FIG. 8. Note
that a description of substantially the same processes in FIG. 8
will be omitted, and only different processes will be
explained.
FIG. 25 shows the input/output characteristics of a printer in
which the maximum value of the 12-bit tone correction table is
"4080". In FIG. 25, broken curve b represents the output density
characteristics, which have been normalized to 10 bits, and its
inverse function (broken curve c) is the tone correction table.
In this case, since a tone correction table having a maximum
value=3886 is obtained using a print head of the middle ejection
amount rank, the output density characteristics indicated by broken
curve b are converted into the characteristics indicated by solid
curve a, and its inverse function (solid curve d) is obtained.
In order to obtain a tone correction table indicated by solid curve
d in step S106 of generating the reverse table, the following two
methods are available.
First Method (b.fwdarw.a.fwdarw.d)
The inverse function of the tone correction table having the
maximum value=3886 is represented by solid curve a. As shown in
FIG. 25, solid curve a sticks to the upper limit in a region where
the input signal value is equal to or higher than 243
(.apprxeq.3886/4080.times.255). Therefore, the range of ordinate y
in FIG. 25 is defined by that from 0 to 1, and curves a and b shown
in FIG. 25 are respectively expressed as y=a(x) and y=b(x) using
function names corresponding to the signs of these curves. Then,
the following relationship is obtained.
The tone correction table is defined by an inverse function y=d(x)
of y=a(x).
Second Method (b.fwdarw.c.fwdarw.d)
y=c(x) as a correction table having a maximum value=4080 is
calculated, and a tone correction table y=d(x) is calculated based
on this table.
Next, x and y in equation (1) used in polynomial approximation of
the reverse table y=d(x) in step S109 of generating the tone
correction table assume real numbers ranging from 0 to 1. Hence, in
order to set the maximum value of the tone correction table to
3886, the calculation result of equation (1) is multiplied by 3886
and the product is rounded to an integer, thus obtaining the tone
correction table.
Tone correction tables for print heads of the large and small
ejection amount ranks can becalculated on the basis of the
calculated tone correction table for a print head of the middle
ejection amount rank by:
Correction value for large ejection amount rank=correction value of
middle ejection amount rank.times.3709/3886
Correction value for small ejection amount rank=correction value of
middle ejection amount rank.times.4080/3886
[Dot Quantity Correction Using Tone Correction Table]
Dot quantity correction using a tone correction table will be
explained below.
Since an ink-jet printer prints an image by directly spraying ink
onto a print medium, if the ink dot quantity (applied quantity) is
larger than the ink acceptable quantity (upper limit value of ink
dot quantity) of a print media (print sheet), beading (ink
overflow, blur due to ink overflow) occurs.
The upper limit value of the ink dot quantity varies depending on
the characteristics of print media. Hence, in order to execute a
color process in correspondence with print media having different
upper limit values of the dot quantity, a tone correction table is
often adjusted with respect to a common color conversion table in
the RGB/CMYK converter 1301 to adjust the dot quantity.
FIG. 26 is a block diagram showing the arrangement of a printer
driver 1300 (color processing unit) which runs on the host, and an
ink-jet printer 1400. FIG. 26 shows the arrangement when the
printer driver makes ejection amount correction of a print head
using a tone correction table. FIG. 27 is a view for explaining the
storage contents of a database 1304 (see FIG. 26).
When color conversion table A and tone correction table A for print
medium A with an ink acceptable quantity of 100% are already
present, for example, color conversion table A for print medium A
is diverted for print medium B with an ink acceptable quantity of
80%, and a tone correction table is re-generated so as to reduce
the size of the database 1304, thus realizing a color process that
meets the ink acceptable amount of print medium B in some
cases.
When R, G, and B image data (300 dpi, 8 bits per color) are input
to a printer driver (color processing unit) 1300, the input image
data undergo RGB.fwdarw.CMYK color conversion by a 3DLUT of a color
converter 1301 after color correction so as to be converted into C,
M, Y, and K data. The C, M, Y, and K image data undergo tone
correction by a tone correction unit 1302, and then multi-valued
pseudo halftoning by a multi-valued error diffusion processor 1303.
Then, the C, M, Y, and K image data are output to a printer 1400.
At this time, color processing tables used in the color converter
1301 and tone correction unit 1302 are selected from a database
1304 on the basis of information associated with the type of print
medium, which is input at, e.g., a user interface (not shown)
provided by the printer driver 1300. For example, when the input
type of print medium corresponds to print medium B, color
conversion table A and tone correction table B shown in FIG. 27 are
set in the color converter 1301 and tone correction unit 1302.
Since tone correction tables A and B having a function of adjusting
the dot quantity are used between print media having different dot
quantities, a color process is done using common color conversion
table A, as shown in FIG. 27. Tone correction table B having a
function of adjusting the dot quantity will be explained below.
Assume that if the ink acceptable quantity of print medium A is
equal to the maximum dot quantity of a color processing table (a
set of a color conversion table and tone correction table)
generated for a print head of the middle ejection amount rank, and
this dot quantity (ink acceptable quantity) is 100%, print medium B
has an ink acceptable quantity 80% of that of print medium A. When
a color conversion table for print medium A is diverted as one of
color conversion table sets for print medium B, a tone correction
table that meets the ink acceptable quantity of print medium B can
be calculated by:
Tone correction value for print medium B=tone correction value for
print medium A.times.0.8
The tone correction table for print medium B sets a dot quantity
80% of that of the tone correction table for print medium A
throughout the medium. Therefore, when tone correction is made
using the calculated tone correction table, image data can undergo
a color process so as not to exceed the ink acceptable quantity of
print medium B even when the color process is done using "color
conversion table for print medium A"+"tone correction table for
print medium B".
[Multi-valued Printer Using Discontinuous Index Patterns]
Along with the price reduction and technical advances of printers,
printers are demanded to assure higher resolution and to have a
smaller memory size in their main bodies. To realize such printers,
an arrangement which uses discontinuous index patterns shown in
FIG. 12 as those of a 600-dpi super pixel of 2.times.2 pixels, and
suppresses the memory size by reducing index numbers is
adopted.
As shown in FIG. 12, four different index patterns, which
respectively have zero, one, two, and four ON dots, are available
as the super pixel, and correspond to 300-dpi, 4-valued data. FIG.
13 summarizes correspondence between the signal values (central
values) and index numbers, as in FIG. 3.
FIG. 14 shows the number of ON dots with respect to the signal
value in a 600-dpi, binary pattern when the index patterns shown in
FIG. 12 are used under the condition shown in FIG. 13. As shown in
FIG. 14, the number of ON dots is expressed by a polygonal line
which inflects at a signal value=170. This is because the number of
ON dots decreases by one and increases by two in a region of
0<signal value<170, while it decreases by two and increases
by four in a region of 170<signal value<255.
Such system in which the number of ON dots increases in a polygonal
line pattern shows the characteristics in which the print density
curve inflects at an injection point of an increase in the number
of ON dots (corresponding to a switching point of index patterns),
as indicated by the solid curve, unlike the print density
characteristics of a multi-valued printer using normal (continuous)
index patterns, which are indicated by the broken curve in FIG. 15.
Therefore, in order to obtain linear print density characteristics
by correction, a tone correction table shown in FIG. 16, which
corresponds to the print density inflection point shown in FIG. 15,
is required.
SUMMARY OF THE INVENTION
The present invention has been made to solve the aforementioned
problems individually or together, and has as its object to
generate a tone correction table which can accurately correct tone
characteristics and can obtain linear output characteristics in a
system in which the output of a print signal is discontinuous with
respect to an input signal.
It is another object of the present invention to generate a tone
correction table which compensates for discontinuous output
characteristics and can set linear output characteristics in a
printing apparatus having discontinuous output characteristics.
It is still another object of the present invention to generate a
tone correction table which can use an identical measurement patch
format independently of printer specifications while reducing the
number of measurement patches to be printed upon generating a tone
correction table, allows processes by simple arithmetic operations,
and can easily attain accurate, faithful halftone reproduction.
In order to achieve the above objects, a preferred embodiment of
the present invention discloses a generation method of generating a
correction table used to correct print characteristics of a
printing apparatus which has nonlinear characteristics between an
input signal and a print signal to be output in accordance with the
input signal, the method comprising steps of:
printing images for measuring amounts of brightness while
compensating for the nonlinear characteristics;
measuring amounts of brightness of the printed images;
generating a tone correction table based on the measured amounts of
brightness; and applying conversion that considers the nonlinear
characteristics to the tone correction table.
A preferred embodiment of the present invention also discloses an
image processing method of generating a correction table used to
correct print characteristics of a printing apparatus which has
nonlinear characteristics between an input signal and a print
signal to be output in accordance with the input signal, the method
comprising steps of:
printing images for measuring amounts of brightness, which include
an image corresponding to an input signal at which the print
characteristics inflect;
measuring amounts of brightness of the printed images; and
generating a tone correction table based on the measured amounts of
brightness.
It is still another object of the present invention to generate a
tone correction table which can compensate for discontinuous output
characteristics to obtain linear output characteristics in a
printing apparatus in which the output of a print signal is
discontinuous with respect to an input signal, and has an arbitrary
maximum value.
It is still another object of the present invention to generate a
tone correction table, which can compensate for discontinuous
output characteristics to obtain linear output characteristics in a
printing apparatus in which the output of a print signal is
discontinuous with respect to an input signal, and has an arbitrary
maximum value, on the basis of a tone correction curve that
compensates for the discontinuous output characteristics.
In order to achieve the above objects, a preferred embodiment of
the present invention discloses a generation method of generating a
correction table used to correct print characteristics of a
printing apparatus which has nonlinear characteristics between an
input signal and a print signal to be output in accordance with the
input signal, the method comprising step of generating, on the
basis of a first tone correction table which is generated while
compensating for the nonlinear characteristics, a second tone
correction table having a maximum value different from the first
tone correction table.
Other features and advantages of the present invention will be
apparent from the following description taken in conjunction with
the accompanying drawings, in which like reference characters
designate the same or similar parts throughout the figures
thereof.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a view showing a layout method of ON dots when 2.times.2
binary pixels of a resolution of 600 dpi are used as one set;
FIG. 2 is a graph showing an average distribution of the number of
index patterns assigned per unit area (ordinate) with respect to
the signal value of an input image signal (abscissa);
FIG. 3 is a table showing the correspondence between the signal
values (central values) and index numbers;
FIG. 4 is a graph showing the average number of 600-dpi, binary ON
dots per unit area (ordinate) with respect to the signal value of
an input image signal (abscissa);
FIG. 5 is a graph showing the print density (ordinate) upon
printing an input signal value (abscissa) as a 600-dpi binary
pattern;
FIG. 6 is a graph showing the tone characteristics which define the
linear relationship between the input signal value and print
density;
FIG. 7 is a graph showing an inverse function table for correcting
the relationship between the input signal value and print
characteristics;
FIG. 8 is a flow chart for explaining a process for deriving a tone
correction table for a 5-valued printer with an input resolution of
300 dpi;
FIG. 9 is a graph showing a tone correction table which allows an
input signal value to pass through it;
FIG. 10 shows measurement patches;
FIG. 11 is a graph showing a "signal value--density" table;
FIG. 12 shows discontinuous index patterns;
FIG. 13 is a table showing the correspondence between the signal
values (central values) and index numbers;
FIG. 14 is a graph showing the number of ON dots with respect to
the signal value in a 600-dpi, binary pattern;
FIG. 15 is a graph showing the print density characteristics of a
system in which the number of ON dots increases in a polygonal line
pattern;
FIG. 16 is a graph showing a tone correction table required for the
system in which the number of ON dots increases in a polygonal line
pattern;
FIG. 17 is a graph for explaining discontinuity that has occurred
in the print density characteristics after tone correction;
FIG. 18 shows a sheet on which 256-level patches are printed;
FIG. 19 is a view showing a method of using sheets on which
256-level patches are printed;
FIG. 20 is a block diagram showing the arrangement of a printer
driver which runs on a host, and an ink-jet printer;
FIG. 21 is a flow chart for explaining the sequence for generating
(deriving) a tone correction table;
FIG. 22 is a graph showing an inverse index table;
FIG. 23 shows measurement patches to which patches of a signal
value corresponding to the switching point of index patterns are
added;
FIG. 24 is a block diagram showing the arrangement of a printer
driver (color processing unit) which runs on a host, and an ink-jet
printer;
FIG. 25 is a graph showing the input/output characteristics of a
printer in which the maximum value of a 12-bit tone correction
table is "4080";
FIG. 26 is a block diagram showing the arrangement of a printer
driver (color processing unit) which runs on a host, and an ink-jet
printer;
FIG. 27 is a table showing the storage contents of a database;
FIGS. 28 and 29 are graphs for explaining the method of generating
a tone correction table;
FIG. 30 is a graph showing a required tone correction table;
FIG. 31 is a flow chart showing the sequence for generating a tone
correction table in the third embodiment;
FIG. 32 is a graph showing a table used to convert a tone
correction table in correction of index components;
FIG. 33 is a flow chart showing the method of generating a tone
correction table in the fourth embodiment;
FIG. 34 is a flow chart showing the method of generating a tone
correction table in the fifth embodiment;
FIG. 35 is a graph for explaining the method of generating a tone
correction table; and
FIG. 36 is a graph for explaining the method of generating a tone
correction table.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Image processes according to preferred embodiments of the present
invention will be described in detail hereinafter with reference to
the accompanying drawings.
First Embodiment
In a multi-valued printer that uses discontinuous index patterns,
when a tone correction table is derived by the same method as in
the prior art, since the tone correction table is generated using
sampled density patches, a table having no inflection point (solid
curve) is obtained unlike a table (broken curve) to be obtained,
thus causing discontinuity in the print density characteristics
after tone correction, as shown in FIG. 17.
When a tone correction table is obtained using sheets, on each of
which 256-level patches are printed, as shown in FIG. 18, for
respective colors, as shown in FIG. 19, the influence of sampling
can be eliminated. However, during printing 1,024 (=256.times.4)
patches, the influences of variations of ejection amount due to the
temperature rise of a print head appear, and the influences of
print errors readily appear on a single sheet and between different
sheets. Under the influences of such errors, the inflection point
of the print density curve deviates, thus causing discontinuity in
the print density characteristics after tone correction.
Furthermore, a smoothing process of a reverse table using a recur
recursive curve requires devices that consider the inflection
point, e.g., use of different recursive curves on the right and
left sides of the inflection point, resulting in a complicated
process.
Hence, this embodiment will explain a method of generating a tone
correction table which can accurately correct tone characteristics
and can obtain linear output characteristics in a system using
discontinuous index patterns.
Also, this embodiment will explain a method of generating a tone
correction table which can compensate for discontinuous output
characteristics and can set linear output characteristics in a
printer using discontinuous index patterns.
Furthermore, this embodiment will explain a method of generating a
tone correction table which can use an identical measurement patch
format independently of printer specifications while reducing the
number of measurement patches to be printed upon generating a tone
correction table, allows processes by simple arithmetic operations,
and can easily attain accurate, faithful halftone reproduction.
The first embodiment will exemplify a case wherein an ink-jet
printer which prints binary data at an output resolution of 600 dpi
uses 300-dpi, 4-valued index patterns shown in FIG. 12, which
correspond to a super pixel of 600-dpi, 2.times.2 pixels, and a
host performs tone correction while considering that ink-jet
printer as a multi-valued (4-valued) printer with an input
resolution of 300 dpi.
[Arrangement]
FIG. 20 is a block diagram showing the arrangement of a printer
driver 300 which runs on a host, and an ink-jet printer 400.
R, G, and B image signals (300 dpi, 8 bits per color) input to a
printer driver 300 are converted into C, M, Y, and K signals (300
dpi, 8 bits per color) by an RGB/CMYK converter 301. Note that a
process for a 300-dpi, 8-bit cyan (C) signal will be explained
below, but the same applies to those for other color component
signals.
The 300-dpi, 8-bit (256-valued) cyan (C) signal undergoes tone
correction in a tone correction unit 302 and is expanded to 12 bits
(4081 values). The expanded 300-dpi, 12-bit C signal undergoes
pseudo halftoning by a multi-valued error diffusion processor 303
to be converted into a 300-dpi, 2-bit (4-valued) multi-valued dot
image.
The 300-dpi, 2-bit multi-valued dot image is transferred (input)
from the host to a printer 400. The 300-dpi, 2-bit multi-valued dot
image input to the printer 400 is rasterized to a 600-dpi, binary
dot image by a dot image rasterize processor 401, which refers to
the index patterns of a super pixel stored in an index pattern
memory 402.
The 600-dpi, binary dot image is stored in a dot image rasterize
buffer 403, and is sequentially sent to a print unit 404 in
correspondence with nozzles arranged in an ink-jet print head at
600-dpi intervals. Print nozzles corresponding to ON dots of the
600-dpi, binary dot image are driven, thus ejecting print liquid
droplets.
Note that the relationship between the input signal value and the
number of ON dots of the printer 400 is expressed by the polygonal
line shown in FIG. 14.
[Generation of Tone Correction Table]
Generation of a tone correction table for the system with the above
characteristics will be explained below.
FIG. 21 is a flow chart showing the sequence for generating
(deriving) a tone correction table. When the host executes software
which is provided as a part of or in association with the printer
driver, a tone correction table can be generated.
Compared to the processes shown in FIG. 8, the steps of generating
an inverse index table (S200), generating an intermediate output
gamma table (S201), and correcting index components (S202) are
added to those shown in FIG. 21, as will be described in detail
later. Note that the same step numbers denote the same processes as
in FIG. 8, and a detailed description thereof will be omitted.
Generation Principle of Tone Correction Table
A tone correction table can be considered as an inverse function of
an "input signal value--output density" function. Therefore, an
output gamma curve is generated using a curve which cancels
discontinuity of index patterns, and the generated output gamma
curve is calculated and converted based on the former curve, thus
obtaining a tone correction table for the discontinuous index
patterns.
A method of generating a tone correction table corresponding to the
discontinuous index patterns will be explained using equations (2)
to (6). Note that a case will be explained wherein the input/output
ranges of all functions are normalized to a range from 0 to 4080
for the sake of simplicity.
where x: a normalized input signal value [0, 4080]
i(.cndot.): a normalized index table (.varies. ink dot
quantity)
input [0, 4080], output [0, 4080]
o(.cndot.): a normalized output density table with respect to the
normalized input signal value
input [0, 4080], output [0, 4080]
n(.cndot.): a normalized output density table with respect to the
input of the normalized dot quantity
input [0, 4080], output [0, 4080]
Note that the normalized input signal value x corresponds to an
image input signal value, and the normalized output density z
corresponds to the output density of the printer. Also, the
normalized index table i(.cndot.) expresses the normalized number
of ON dots (ink dot quantity) with respect to the input signal
value when the index patterns of this embodiment are used. Upon
calculating values i(x) over the full definition range [0, 4080] of
the normalized input signal value x, a graph shown in FIG. 14 is
obtained.
The normalized output density table o(.cndot.) with respect to the
normalized input signal value x expresses the normalized output
density with respect to the normalized input signal value when an
output is made without tone correction, i.e., when an output is
made while the relationship between the number of ON dots and input
signal value is, as shown in FIG. 14. The normalized output density
table o(.cndot.) with respect to the normalized input signal value
is as shown in FIG. 15 due to the presence of the inflection point
(switching point of slopes) shown in FIG. 14.
The normalized output density table n(.cndot.) with respect to the
normalized input signal value x expresses the normalized output
density with respect to the normalized input signal value when the
number of ON dots and the input signal value have a proportional
relationship, as shown in FIG. 6. The normalized output density
table n(.cndot.) with respect to the normalized input signal value
x in this case is as shown in FIG. 5.
In this embodiment, the output density characteristics without any
tone correction are given by equation (2), and are expressed as a
graph, as indicated by the solid curve in FIG. 15. Therefore, a
tone correction table to be obtained in this embodiment is given by
equation (3) as an inverse function of equation (2), and is
expressed by the bold solid curve in FIG. 16.
The conventional method of generating a tone correction table aims
at a system having continuous print characteristics shown in FIG.
6. Therefore, that method can be applied to a system having output
density characteristics shown in FIG. 5. This corresponds to:
The output density characteristics (FIG. 5) of the system which
uses continuous index patterns and corresponds to equation (7)
represent the output density with respect to the input image signal
value. At the same time, such characteristics also represent the
output density with respect to the number of ON dots at 600 dpi. In
the system of this embodiment, the number of ON dots with respect
to the input signal value can be expressed by i(x), and has
characteristics shown in FIG. 14.
Hence, the output density characteristics of this embodiment can be
expressed by equation (4) using the output density characteristics
n(.cndot.) of the system using continuous index patterns, and the
normalized index table i(.cndot.). By modifying equation (4), a
tone correction table is derived.
Using the conventional method of generating a tone correction
table, n.sup.-1 (.cndot.) indicated by the solid curve in FIG. 7 is
calculated, and is given to both sides of equation (4) to yield
equation (5). An inverse function i.sup.-1 (.cndot.) of i(.cndot.)
is obtained, as indicated by the broken curve in FIG. 22, and is
given to both sides of equation (5) to yield equation (6).
On the other hand, as can be seen from equation (3) as a
conventional tone correction table, the tone correction table can
be defined to have the normalized output density z as an input and
the normalized input signal value x as an output. Upon examining
equation (6) based on this fact, i.sup.-1 (n.sup.- (.cndot.))
defines a tone correction table of this embodiment.
In order to calculate the output density characteristics n(.cndot.)
of the system using continuous index patterns in the system using
discontinuous index patterns like in this embodiment, patches may
be output using the inverse function i.sup.-1 (.cndot.) of
i(.cndot.) in place of the tone correction table, and the densities
of the output patches may be measured.
The processes unique to this embodiment shown in the flow chart of
FIG. 21 will be explained below.
Inverse Index Table (S200)
An inverse index table indicated by the solid curve in FIG. 22 is
an inverse function (broken curve) of a table of the number of ON
dots with respect to the input signal value when discontinuous
index patterns are used. As shown in FIG. 22, since this table can
be easily generated, a detailed description of the generation
method will be omitted.
That is, in order to obtain the output density characteristics
n(.cndot.) of the system using continuous index patterns, an
inverse index table is generated (S200), and is set as a tone
correction table upon outputting measurement patches (S102).
By outputting patches using the inverse index table, the output of
the number of ON dots, which is linear to the input image signal
value, as shown in FIG. 6, can be obtained after
multi-valued.fwdarw.binary conversion (rasterize index patterns) by
the printer 400. In other words, using the inverse index table, the
number of ON dots can be linear to the input image signal value,
and even when patches (FIG. 10) sampled at appropriate intervals
are used, a sampling error of a density characteristic switching
point shown in FIG. 17 can be prevented.
If sampled patches can be used, highly reliable print density
characteristics can be obtained while suppressing the influences of
variations of the ejection amount due to the temperature rise of
the print head even in a system in which the density reproduction
characteristics on a single sheet or between different sheets are
unstable. Of course, many processes in the conventional generation
method of a tone correction table can be used in generation of a
tone correction table in a system with nonlinear print
characteristics like in the printer 400 of this embodiment. More
specifically, the same processes as those in the conventional
generation method of a tone correction table are executed in steps
S102 to S108 in FIG. 21.
Generation of Intermediate Output Gamma Table (S201)
Both the input and output of the input/output range [0, 1] of the
reverse table finely adjusted by the process in step S108 are
adjusted to [0, 4080], thus obtaining n.sup.-1 (.cndot.) called an
"intermediate output gamma table" in this embodiment.
Correction of Index Components (S203)
As can be seen from comparison between equations (3) and (6), a
tone correction table o.sup.-1 (.cndot.) to be obtained is
equivalent to i.sup.-1 (n.sup.-1 (.cndot.)), as described above,
Hence, by synthesizing i.sup.-1 (.cndot.) and n.sup.-1 (.cndot.),
o.sup.-1 (.cndot.) is obtained.
That is, output values are examined by referring to n.sup.-1
(.cndot.) shown in FIG. 7 for all input values within the input
range [0, 255]. These output values are input to the inverse index
table i.sup.-1 (.cndot.) indicated by the solid curve in FIG. 22,
thus obtaining the synthesis output values i.sup.-1 (n.sup.-1
(.cndot.)) of the table. In other words, in step S202 the tone
correction table n.sup.-1 (.cndot.) having no inflection point
shown in FIG. 7 is mathematically converted to calculate the tone
correction table with the inflection point, indicated by the bold
solid curve in FIG. 16.
By calculating the discontinuous inflection point of the tone
correction table by mathematical conversion, "deviation of the
inflection point" explained as a problem can be prevented.
The synthetic output values i.sup.-1 (n.sup.-1 (.cndot.)) for
respective input image signal values calculated in the
aforementioned sequence are converted into a table, and that table
is substituted in a sequence for the tone correction table o.sup.-1
(.cndot.), thus obtaining a tone correction table (output gamma
table) indicated by the bold solid curve in FIG. 16 (S109).
Note that the obtained tone correction table is set in the tone
correction unit 302 of the printer driver 300.
Second Embodiment
An image processing apparatus according to the second embodiment of
the present invention will be described below. Note that the same
reference numerals in this embodiment denote the same parts as
those in the first embodiment, and a detailed description thereof
will be omitted.
The processing sequence in the second embodiment is substantially
the same as that in FIG. 8, and differences will be explained
below.
Set Measurement Patches (S103)
When the sampled patches shown in FIG. 10 are used, a problem is
posed at the inflection point of the tone correction table which is
caused by switching of index patterns. To solve this problem, the
second embodiment sets patches which include those of a signal
value corresponding to the index pattern switching point as
measurement patches, as shown in FIG. 23.
Smoothing of Reverse Table Using Recursive Curve (S107)
In the second embodiment, the output density plots always include
the index pattern switching point since the patches of a signal
value corresponding to the index pattern switching point are added.
However, if the reverse table is smoothed using the recursive curve
by the conventional method, this switching point may not be
smoothed. Hence, in the second embodiment, the output density plots
are divided into regions on the right and left sides of the
switching point to set recursive curves, thereby smoothing the
reverse table.
Fine Adjustment of Reverse Table (S108)
The reverse table is finely adjusted to adjust any deviation due to
oscillations of the recursive curve and that of the end point of
the tone correction table.
In the second embodiment, since the recursive curves are set by
dividing the output density plots into two regions to smooth the
reverse table, another deviation occurs at the index pattern
switching point as the boundary of the two recursive curves.
Therefore, in the second embodiment, a deviation of the switching
point as the boundary of the two recursive curves is adjusted in
addition to the aforementioned adjustment of deviations.
As described above, according to the above embodiments, a tone
correction table which can accurately correct tone characteristics
and can obtain linear output characteristics in a system in which
the output of a print medium is discontinuous with respect to an
input signal value can be generated.
Also, the method of generating a tone correction table which can
compensate for discontinuous output characteristics to obtain
linear output characteristics in a printing apparatus having
discontinuous output characteristics can be provided.
Furthermore, the method of generating a tone correction table which
can use an identical measurement patch format independently of
printer specifications while reducing the number of measurement
patches to be printed upon generating a tone correction table,
allows processes by simple arithmetic operations, and can easily
attain accurate, faithful halftone reproduction can be
provided.
The above embodiments adopt the method of smoothing data using a
recursive curve after an inverse table of an input signal
value--output density table is calculated unlike in Japanese Patent
Publication No. 8-2659. However, a tone correction table may be
obtained using a method of calculating an inverse table after data
of an input signal value--output density table are smoothed, as in
Japanese Patent Publication No. 8-2659.
In the above example, combinations of dots which have the same size
and correspond to an identical ink type are used as index patterns.
However, in case of a printer which can selectively print small and
large dots, combinations of small and large dots may be used as
index patterns. Also, in case of a printer which uses two types of
inks, i.e., light and dark inks, combinations of light and dark
dots may be used as index patterns. In these cases, the method of
generating a tone correction table of the above embodiment can be
applied.
Furthermore, in the above example, the number of inflection points
of the print density curve caused by switching of index patterns is
one. However, the number of inflection points is not limited to
one, and the method of generating a tone correction table of the
above embodiment can be applied to a system having two or more
inflection points.
Of course, the method of generating a correction table of this
embodiment is not limited to an ink-jet printer using discontinuous
index patterns, and can be widely applied to various other printing
apparatuses which have discontinuous output characteristics with
respect to input characteristics (e.g., a CRT hard copy
apparatus).
Third Embodiment
An image processing apparatus according to the third embodiment of
the present invention will be described below. Note that the same
reference numerals in this embodiment denote the same parts as
those in the first embodiment, and a detailed description thereof
will be omitted.
[Problem in Tone Correction Table Generation By Above Method]
When a tone correction table whose maximum value corresponds to
point P in FIG. 28 and is smaller than a full-range value (4080 in,
e.g., a table of a 12-bit output) is to be generated by the method
shown in FIG. 21 for a multi-valued printer using discontinuous
index patterns, a tone correction table which has a desired maximum
value (point P in FIG. 28) and can define linear output density
characteristics after tone correction cannot be obtained.
More specifically, a full-range tone correction table is indicated
by broken curve C1 in FIG. 29, while a tone correction table which
is obtained by a method of deriving a tone correction table having
an arbitrary maximum value for a printing apparatus having linear
output characteristics, and has a desired maximum value is
indicated by two-dashed chain curve C0. Hence, a tone correction
table which compensates for nonlinearity of the printer to obtain
linear output density characteristics cannot be obtained.
When a tone correction table having a desired maximum value is
generated by a conversion method of a tone correction table by,
e.g., simple variable scaling on the basis of a tone correction
table corresponding to nonlinearity of a printer, a tone correction
table shown in FIG. 28 is obtained. Hence, a required tone
correction table shown in FIG. 30 cannot be obtained, and the tone
characteristics after tone correction are worsened.
More specifically, tone correction tables for the large and small
ejection amount ranks, which are obtained by the aforementioned
method on the basis of a tone correction table for a print head of
the middle ejection amount rank shown in FIG. 28, are as shown in
FIG. 28, and the inflection points of the tone correction tables
used to compensate for the printer output characteristics are
located on the extension of point Q on the x-axis. However, when a
printer having the output characteristics shown in FIG. 14 makes
output, the inflection points of the tone correction tables for the
large and small ejection amount ranks must be located on the
extension of point P on the y-axis, as shown in FIG. 30, so as to
obtain linear output density characteristics.
Furthermore, in the tone correction table generation method for the
aforementioned system which uses discontinuous index patterns and
has nonlinear output characteristics, when an intermediate output
gamma table undergoes maximum value adjustment and then index
component correction to convert a tone correction table indicated
by two-dashed chain curve C0 in FIG. 29, a tone correction table
indicated by solid curve C2 is obtained. However, tone correction
table C2 has a maximum value different from the desired maximum
value.
[Solutions]
The following three solutions may be used to solve the
aforementioned problem:
(1) Tone correction tables are generated on the basis of the output
density data of a printer for respective tone correction tables
having different maximum values.
(2) Tone correction tables having different maximum values are
obtained by analytic conversion on the basis of one intermediate
output gamma table.
(3) Conversion is made based on one tone correction table to
analytically obtain a tone correction table having an arbitrary
maximum value.
These three solutions will be described below in turn as the third
embodiment.
The third embodiment will exemplify a case wherein an ink-jet
printer which prints binary data at an output resolution of 600 dpi
uses 300-dpi, 4-valued index patterns shown in FIG. 12, which
correspond to a super pixel of 600-dpi, 2.times.2 pixels, and a
host performs tone correction while considering that ink-jet
printer as a multi-valued (4-valued) printer with an input
resolution of 300 dpi.
Note that the arrangement of a printer driver 1300 (color
processing unit) which runs on the host and the ink-jet printer
1400 are the same as that shown in FIG. 24, and a detailed
description thereof will be omitted. Also, the relationship between
the input image signal value and the number of ON dots of the
printer 1400 is expressed by a nonlinear polygonal line, as shown
in FIG. 14.
A method of generating a tone correction table for a system having
the output characteristics shown in FIG. 12 will be described
below.
FIG. 31 is a flow chart showing the sequence for generating
(deriving) a tone correction table of the third embodiment. When
the host executes software which is provided as a part of or in
association with the printer driver, a tone correction table can be
generated.
Compared to the processes shown in FIG. 21, the step of setting a
maximum value of a tone correction table (S301) is added to those
shown in FIG. 21, as will be described in detail later. Note that
the same step numbers denote the same processes as in FIGS. 8 and
21, and a detailed description thereof will be omitted.
Setup of Maximum Value of Tone Correction Table (S301).
The maximum value of a tone correction table to be obtained is set.
In this case, since a tone correction table for a print head of the
middle ejection amount rank is to be obtained, "3886" (see FIG. 28)
is set as the maximum value of the correction table.
Setup of Output Gamma for Outputting Measurement Patches (S102)
In order to calculate the output density characteristics n(.cndot.)
in a system using continuous index patterns using equation (1), an
inverse index table generated in step S200 is set as a tone
correction table used to output measurement patches. By outputting
patches using the inverse index table, the output of the number of
ON dots, which is linear to the input image signal value, as shown
in FIG. 6, can be obtained after multi-valued.fwdarw.binary
conversion (rasterize index patterns) by the printer 1400.
By printing patches using the inverse index table in this way, the
number of ON dots can be linear to the input image signal value,
and even when patches (FIG. 10) sampled at appropriate intervals
are used, a sampling error of a density characteristic switching
point shown in FIG. 17 can be prevented. Since arbitrary sampling
is allowed, highly reliable print density characteristics can be
obtained even in a system in which the density reproduction
characteristics on a single sheet or between different sheets are
unstable like an in-jet printer. Furthermore, many processes in the
conventional generation method can be used upon generating a tone
correction table in a system with nonlinear print
characteristics.
Generation of Intermediate Output Gamma Table (S201)
The input and output of the input/output range [0, 1] of the
reverse table finely adjusted by the process in step S108 are
respectively adjusted to [0, 4080] and [0, 3886].
Correction of Index Components (S202)
As in the above description, a process for adding nonlinearity of
the printer input/output characteristics to the intermediate output
gamma table obtained by removing the nonlinearity of the printer
input/output characteristics is executed. A table used in
conversion of a tone correction table in index component correction
as a characteristic feature of the third embodiment will be
explained below with reference to FIG. 32.
An intermediate output gamma table indicated by the solid curve in
FIG. 7 may be converted using an inverse index table which has a
maximum value=4080, and is indicated by the one-dashed chain curve
in FIG. 32. More specifically, all input values 0 to 255 are
corrected using the intermediate output gamma table, and output
values of the inverse index table are calculated for the obtained
corrected input values, thus attaining conversion. However, when
the intermediate output gamma table whose maximum value is not 4080
is converted in this way, a maximum value portion has obviously
deviated. This is because when the maximum value is 3886 as in the
third embodiment, the correction value to be input to the inverse
index table is present in regions X and Y shown in FIG. 32 but is
not present in region Z.
In consideration of the above problems, the third embodiment uses a
table which is obtained based on the following conditions and
changes depending on the maximum value of a tone correction table,
as indicated by the solid curve in FIG. 32, as an inverse index
table used to correct index components.
(a) An inflection between regions X and Y caused by the printer
input/output characteristics is not influenced by a change in
maximum value of a tone correction table.
(b) The maximum value of an intermediate output gamma table is
preserved by conversion using an inverse index table. In other
words, this value remains the same before and after conversion.
Therefore, correction values of input values 0 to 255 are
respectively calculated using an intermediate output gamma table as
correction values of a tone correction table, and correction using
an inverse index table is made using the calculated correction
values as inputs, thus obtaining the correction values of the
inverse index table.
Generation of Tone Correction Table (S109)
The intermediate output gamma table converted by the inverse index
table in step S202 is converted into a format that can be stored in
a tone correction table database 1305 shown in FIG. 24, thus
generating a tone correction table.
In this manner, the tone correction table for the print head of the
middle ejection amount rank can be obtained.
In the above description, the method of deriving the tone
correction table for the print head of the middle ejection amount
rank has been explained. Also, tone correction tables for print
heads of the large and small ejection amount ranks can be generated
in the same sequence. More specifically, upon generating tone
correction tables for print heads of the large and small ejection
amount ranks, the maximum values are respectively set to 3709 and
4080, thus obtaining tone correction tables compatible to ejection
amount correction.
Fourth Embodiment
A method of generating a tone correction table according to the
fourth embodiment of the present invention will be described below.
Note that the same reference numerals in this embodiment denote the
same parts as those in the first and third embodiments, and a
detailed description thereof will be omitted.
A case will be explained below wherein a color process is performed
in a system using a printer 1400 which has the same ejection amount
correction function as in the third embodiment and has nonlinear
output characteristics, as shown in FIG. 24.
The fourth embodiment calculates an intermediate output gamma table
while compensating for the nonlinear output characteristics of the
printer 1400 using the method described in, e.g., the third
embodiment upon calculating a tone correction table of the printer
1400 with the nonlinear output characteristics, and a tone
correction table having an arbitrary maximum value is calculated
using the obtained intermediate output gamma table. More
specifically, an intermediate output gamma table having a maximum
value=3886 is converted to generate an intermediate output gamma
table having an arbitrary maximum value (e.g., 4706), and the
intermediate output gamma table having the maximum value=4706 is
converted into a tone correction table in a format that considers
the nonlinear output characteristics of the printer 1400.
FIG. 33 is a flow chart showing the method of generating a tone
correction table of the fourth embodiment. Note that a case will be
described below wherein a tone correction table for a print head of
the large ejection amount rank, which has a maximum value=3709, is
generated using an intermediate output gamma table having a maximum
value=3886 for a print head of the middle ejection amount rank,
which is generated by the same method as in the third
embodiment.
Setup of Intermediate Output Gamma (S501)
An intermediate output gamma table which is generated by the same
method as in the third embodiment and has a maximum value=3886 is
set. Note that this intermediate output gamma table has no
inflection point, as in curve C1 in FIG. 36 (the maximum value is
different), since it is obtained while compensating for the
nonlinear output characteristics of the printer.
Setup of Maximum value of Output Gamma (S502)
A maximum value of a tone correction table to be obtained is set.
In the fourth embodiment, a maximum value--3709 is set for a print
head of the large ejection amount rank.
Generation of Inverse Index Table (S200)
A table used to convert the intermediate output gamma table
obtained while compensating for the nonlinear output
characteristics of the printer into a tone correction table
corresponding to the nonlinear output characteristics of the
printer is generated. This process is substantially the same as
"generation of inverse index table (S200)" in the third embodiment,
except that the maximum value of the converted tone correction
table is 3709. Therefore, the table generated in this step is
obtained by replacing an input signal value "3886" in FIG. 32 by
"3709". Of course, the inverse index table generated in this step
is used in correction of index components (S202).
Decimation Process of Output Gamma (S503)
Conversion is made on the basis of the tone correction table having
the arbitrary maximum value (3886 in this case), which is set in
the step of setting the intermediate output gamma (S501), thereby
generating an intermediate output gamma table having an arbitrary
maximum value (3709 in this case). An arbitrary method can be used
as the conversion method between tone correction tables having
different maximum values. For example, the output values of the
tone correction table with the set maximum value can undergo simple
variable scaling. A tone correction table after variable scaling
upon execution of simple variable scaling is given by:
where ROUND(.cndot.) is a function which inputs a solid number as
an argument, rounds off the input solid number, and converts that
value into an integer.
Correction of Index Components (S202)
The same process as in the third embodiment is executed using the
inverse index table generated in step S200.
Generation of Tone Correction Table (S109)
The converted table is converted into a format of a tone correction
table as in the third embodiment.
In the above description, the method of generating the tone
correction table for the print head of the middle ejection amount
rank has been explained. Also, tone correction tables for print
heads of the large and small ejection amount ranks can be generated
in the same sequence.
In the above description, a tone correction table for a print head
of another ejection amount rank is calculated on the basis of the
already generated tone correction table for a print head of the
middle ejection amount rank. However, tables for any ejection
amount ranks may be used a base tone correction table.
Fifth Embodiment
A method of generating a tone correction table according to the
fifth embodiment of the present invention will be described below.
Note that the same reference numerals in this embodiment denote the
same parts as those in the first and third embodiments, and a
detailed description thereof will be omitted.
The fifth embodiment will exemplify a case wherein a tone
correction table having a maximum value=3709 for a print head of
the large ejection amount rank is generated on the basis of a tone
correction table generated using a print head of the middle
ejection amount rank in a system using a printer which has the same
ejection amount correction function as in the third embodiment and
has nonlinear output characteristics, as shown in FIG. 24.
FIG. 34 is a flow chart showing the method of generating a tone
correction table of the fifth embodiment.
Setup of Output Gamma (S600)
A tone correction table which is generated by the same method as in
the third embodiment and has a maximum value=3886 is set.
Setup of Maximum value of Output Gamma (S502)
A maximum value of a tone correction table is set to 3709 for a
print head of the large ejection amount rank.
Generation of Index Table (S601)
A table which is indicated by the solid curve in FIG. 35 and is an
inverse function (broken curve) of the number of ON dots with
respect to the input signal value upon using discontinuous index
patterns, is generated. Since this table can be generated easily, a
description of its generation method will be omitted.
Generation of Inverse Index Table (S200)
A table used to convert an intermediate output gamma table obtained
while compensating for the nonlinear output characteristics of the
printer into a tone correction table corresponding to the nonlinear
output characteristics of the printer is generated. This process is
substantially the same as "generation of inverse index table
(S200)" in the third embodiment, except that the maximum value of
the converted tone correction table is 3709. Therefore, the table
generated in this step is obtained by replacing an input signal
value "3886" in FIG. 32 by "3709".
Generation of Intermediate Output Gamma (S201)
The output gamma table, which is set in step S600, corresponds to
the nonlinear characteristics of the printer, and has the maximum
value=3886, is converted using the index table generated in step
S601 into a table which is obtained while compensating for the
nonlinearity of the printer, as in the intermediate output gamma
table of the third embodiment. More specifically, correction values
of the output gamma table are calculated for all inputs, conversion
is made by the index table using these correction values as inputs,
and output values are used as an intermediate output gamma table
(see the following equation).
Intermediate Opg[i]=index table [Opg[i]] where Opg is the
abbreviation for "output gamma", and i is an integer ranging from 0
to 255.
Decimation Process of Output Gamma (S503)
The obtained intermediate output gamma table undergoes a decimation
process as in the fourth embodiment.
Correction of Index Components (S202)
The same process as in the fourth embodiment is executed using the
inverse index table generated in step S200.
Generation of Tone Correction Table (S109)
The converted table is converted into a format of a tone correction
table as in the third embodiment.
The above embodiments adopt the method of smoothing data using a
recursive curve after an inverse table of an input signal
value--output density table is calculated unlike in Japanese Patent
Publication No. 8-2659. However, a tone correction table may be
obtained using a method of calculating an inverse table after data
of an input signal value--output density table are smoothed, as in
Japanese Patent Publication No. 8-2659.
In the above example, combinations of dots which have the same size
and correspond to an identical ink type are used as index patterns.
However, in case of a printer which can selectively print small and
large dots, combinations of small and large dots may be used as
index patterns. Also, in case of a printer which uses two types of
inks, i.e., light and dark inks, combinations of light and dark
dots may be used as index patterns. In these cases, the method of
generating a tone correction table of the above embodiment can be
applied.
Furthermore, in the above example, the number of inflection points
of the print density curve caused by switching of index patterns is
one. However, the number of inflection points is not limited to
one, and the method of generating a tone correction table of the
above embodiment can be applied to a system having two or more
inflection points.
Of course, the method of generating a correction table of this
embodiment is not limited to an ink-jet printer using discontinuous
index patterns, and can be widely applied to various other printing
apparatuses which have discontinuous output characteristics with
respect to input characteristics (e.g., a CRT hard copy
apparatus).
Modification
Simple Method of Index Component Correction
In the fourth and fifth embodiments, a table which preserves a
desired maximum value, as shown in FIG. 32, is used as an inverse
index table used in correction of index components. However, an
inverse index table that preserves a maximum value=4080 shown in
FIG. 35 may be used in place of the table which preserves a desired
maximum value, the terminal end portion of a tone correction table
(curve C1 shown in FIG. 36) that has undergone index component
correction using the inverse index table may be interpolated by a
line to obtain curve C2 shown in FIG. 36.
In this manner, although the accuracy drops slightly, one permanent
table can be used as an inverse index table. In this case, a
simpler system arrangement can be realized since the memory
required to store the inverse index table can be reduced, the
processing time required to re-calculate the inverse index table
can be shortened, and so forth.
Another Method of Output Gamma Decimation Process
In the above embodiments, as a method of generating tone correction
tables having different maximum values, the tone correction values
of the intermediate output gamma table, which is generated while
compensating for the nonlinear output characteristics, undergo
simple variable scaling to obtain tone correction tables having
different maximum values. When all tone correction tables for
respective colors undergo simple variable scaling, the color
balance after the output gamma decimation process can be maintained
even using tone correction tables having different maximum values.
Note that a method that can preserve the linearity of the output
density characteristics may be used as the output gamma decimation
process. In this case, a color process can be done using tone
correction tables having different maximum values, and the density
characteristics of an output image can be preserved.
Arrangement of System Using Tone Correction Table
The tone correction table generated in each of the above
embodiments is used in a system which makes ejection amount
correction of a print head. However, the generated tone correction
table can also be used in a system which makes dot quantity
correction shown in FIG. 26.
Note that tone correction tables corresponding to respective
ejection amount ranks of print heads are as shown in FIG. 30, and
are not simply integer multiples of the full definition range among
them. That is, even when the maximum value of a tone correction
table is converted using a dot quantity ratio, a target dot
quantity cannot often be achieved.
When the generated tone correction table is used in dot quantity
correction, for example, a measure for detecting the maximum value
of a tone correction table that meets a dot quantity while
gradually decreasing the maximum value of a tone correction table
is required.
<Other Embodiment>
The present invention can be applied to a system constituted by a
plurality of devices (e.g., host computer, interface, reader,
printer) or to an apparatus comprising a single device (e.g.,
copying machine, facsimile machine).
Further, the object of the present invention can also be achieved
by providing a storage medium storing program codes for performing
the aforesaid processes to a computer system or apparatus (e.g., a
personal computer), reading the program codes, by a CPU or MPU of
the computer system or apparatus, from the storage medium, then
executing the program.
In this case, the program codes read from the storage medium
realize the functions according to the embodiments, and the storage
medium storing the program codes constitutes the invention.
Further, the storage medium, such as a floppy disk, a hard disk, an
optical disk, a magneto-optical disk, CD-ROM, CD-R, a magnetic
tape, a non-volatile type memory card, and ROM can be used for
providing the program codes.
Furthermore, besides aforesaid functions according to the above
embodiments are realized by executing the program codes which are
read by a computer, the present invention includes a case where an
OS (operating system) or the like working on the computer performs
a part or entire processes in accordance with designations of the
program codes and realizes functions according to the above
embodiments.
Furthermore, the present invention also includes a case where,
after the program codes read from the storage medium are written in
a function expansion card which is inserted into the computer or in
a memory provided in a function expansion unit which is connected
to the computer, CPU or the like contained in the function
expansion card or unit performs a part or entire process in
accordance with designations of the program codes and realizes
functions of the above embodiments.
In a case where the present invention is applied to the aforesaid
storage medium, the storage medium stores program codes
corresponding to the flowcharts described in the embodiments.
As many apparently widely different embodiments of the present
invention can be made without departing from the spirit and scope
thereof, it is to be understood that the invention is not limited
to the specific embodiments thereof except as defined in the
claims.
* * * * *